Indonesia Fires & Floods

Danielle Robinson & Julia Bowling
December 21, 2015

Agenda

  • Background
  • Disasters
  • Timeline
  • Geography and Environment
  • Sources & Variables
  • Analysis
  • Discussion
  • Policy Recommendation

Disasters in Indonesia

  • Floods, Earthquake, Landslides, Volcano, Epidemic, Forest Fires, Storms, Drought, Tsunami, Strong Wind, Floods and Landslides, Surge, Transportation Accident, Forest Fire, Eruption, Conflict, Earthquake and Tsunami, Terrorism, Industrial Accident, Plague, Climate Change, Hunger

Natural Disasters: Loss & Distribution

pie_chart

pie_chart_2

Disasters in Indonesia: Our Focus

  • Fires
  • Floods
  • Forest Fires

Disasters in Indonesia

  • Timeline
  • Major Events
  • Why Indonesia?
  • Factor: Geography

Floods & Fires: A National Timeline

  • 1988 - NASA climate scientist James Hansel declared global warming as a cause of increased heatwaves and drought
  • 1995 - Larson A arctic ice shelf collapse in 1995 (Source: Greenpeace)
  • 2004 - Tsunami (Aceh)
  • 2006 - Sidoarjo mudslide, 40,000 evacuated (East Java)
  • 2010 - Wasior floods, 145 dead, 7,900 evacuated (West Papau)
  • 2013 - Greenhouse gases measured above 400 ppm (Manau Loa)
  • 2014 - Karangkobar landslide, approx 100 deaths (Central Java)
  • 2014 - Flood, 80,000 displaced (North Sulawesi)
  • 2015 - Great Garuda construction begins (Jakarta)

Background of Indonesia

languages

  • Population over 255 million
  • Over 17,000 islands
  • 700 regional languages
  • Industry and Agriculture
  • Cost of air pollution from CO2 emissions: $400 million

Indonesian Geography

  • Provinces & Islands Indonesia

Java

greatgaruda

  • 58% of Indonesian population
  • Jakarta: Sinking City
  • Great Garuda: Seawall

Sumatra

  • Aceh 2004 Tsunami elephant

Kalimantan (Borneo)

forestfires

  • Forest Fires
  • Peatlands
  • Government Incentive
  • Community Adaptation Plans

Sulawesi

sulawesi

  • 2014 flood in north Sulawesi
  • 80,000 evacuees
  • $580 million in damages

Environmental Factors

peatland

  • Peatlands
  • Primary Forest
  • Aforestation and destruction of habitat
  • Endemic Species, 36%

Environmental Challenges

  • Flooding
  • Depleted Aquifers
  • Sea Level Rise
  • Forest Fires
  • Intentional Burning

Flooding

rescue

  • Highest incidence of natural disasters each year
  • Lack solid waste management system
  • Low capacity for water purification
  • High levels of E. Coli

Depleted Aquifers

  • Illegal groundwater extraction
  • Shallow wells
  • Subsidence
  • 40% of Jakarta below sea level

Sea Level Rise (SLR)

  • Rise in global temperatures
  • Global Annual Average SLR: 0.33 inches
  • Jakarta: 3 inches

Forest Fires

  • Palm Oil and Pulp-Paper concessions
  • Strong Winds; Drought; Drained Peatlands; CO2 Emissions forest_conservation

Multi-Lateral Frameworks

  • Hyogo Framework for Action 2005 - 2015
  • Sendai Framework for Disaster Risk Reduction 2015 - 2030
  • Bali Action Plan 2007

Data

  • World Bank
  • DesInventar (UN International Strategy for Disaster Risk Reduction)
  • 4 Islands: Sulawesi, Sumatra, Java, Kalimantan
  • 26 Provinces
  • 4488 events
  • Dates 1990 - 2012

Questions

  • What type of natural disaster, fires or floods, contributes more to loss of livelihood and economic loss?
  • Does conservation of forests help to decrease economic loss?
  • Does the increased occurance of natural disasters increase economic loss?

Hypotheses

  • H0: There is no difference in economic loss from natural disasters including fires, forest fires, and floods.
  • HA: There is an increase in economic loss from natural disasters including fires, forest fires, and floods.

Assumptions

  • 1. Increase in injured people after floods
  • 2. Higher fire risk between May and September
  • 3. Increase in local government intervention, decrease in fire loss
  • 4. Inverse relationship between mobile technology use and number of people injured or deaths
  • 5. Dredged areas experience less loss
  • 6. Floods increase with high levels of groundwater extraction
  • 7. Forest Fire risk increases in older concessions

Data Work

library("ggplot2")
library("ggplot2")
library("tidyr")
library("dplyr")
require(scales)
opts_chunk$set(fig.width=5, fig.height=5)
indo <- read.csv("newindo.csv")
code <- read.csv("islands.csv")
indo <- left_join(indo, code, by=NULL, copy=FALSE)

Data Work, continued.

indo.agg <- indo %>%
  group_by(Event, Province, Year, Island) %>%
  summarise(events=sum(DataCards, na.rm=TRUE), deaths=sum(Deaths, na.rm=TRUE), injured=sum(Injured, na.rm=TRUE), missing=sum(Missing, na.rm=TRUE), housesdestroyed=sum(Houses.Destroyed, na.rm=TRUE), housesdamaged=sum(Houses.Damaged, na.rm=TRUE), victims=sum(Victims, na.rm=TRUE), affected=sum(Affected, na.rm=TRUE), evacuated=sum(Evacuated, na.rm=TRUE), lossUSD=sum(Losses..USD, na.rm=TRUE), lossLOCAL=sum(Losses..Local, na.rm=TRUE), cropdamage=sum(Damages.in.crops.Ha., na.rm=TRUE), agriculture=sum(Agriculture, na.rm=TRUE), power=sum(Power.and.Energy, na.rm=TRUE), relief=sum(Relief, na.rm=TRUE), forest=sum(Plantation.Forest, na.rm=TRUE))

Final Dataset

summary(indo.agg)

         Event                   Province        Year     
 FIRE       :162   KALIMANTAN TIMUR  : 31   Min.   :1990  
 FLOODS     :290   JAWA TENGAH       : 28   1st Qu.:2003  
 FOREST FIRE: 41   SULAWESI SELATAN  : 27   Median :2006  
                   JAWA BARAT        : 26   Mean   :2006  
                   SUMATERA BARAT    : 26   3rd Qu.:2010  
                   KALIMANTAN SELATAN: 25   Max.   :2012  
                   (Other)           :330                 
        Island        events           deaths           injured       
 Java      :129   Min.   :  1.00   Min.   :  0.000   Min.   :    0.0  
 Kalimantan: 98   1st Qu.:  2.00   1st Qu.:  0.000   1st Qu.:    0.0  
 Sulawesi  : 90   Median :  5.00   Median :  0.000   Median :    0.0  
 Sumatra   :176   Mean   : 12.35   Mean   :  5.032   Mean   :  412.6  
                  3rd Qu.: 12.00   3rd Qu.:  4.000   3rd Qu.:    3.0  
                  Max.   :183.00   Max.   :347.000   Max.   :60149.0  

    missing        housesdestroyed   housesdamaged        victims 
 Min.   :  0.000   Min.   :    0.0   Min.   :    0.0   Min.   :0  
 1st Qu.:  0.000   1st Qu.:    0.0   1st Qu.:    0.0   1st Qu.:0  
 Median :  0.000   Median :    0.0   Median :    0.0   Median :0  
 Mean   :  1.213   Mean   :  279.2   Mean   :  341.7   Mean   :0  
 3rd Qu.:  0.000   3rd Qu.:   77.0   3rd Qu.:    0.0   3rd Qu.:0  
 Max.   :140.000   Max.   :21924.0   Max.   :26623.0   Max.   :0  

    affected         evacuated         lossUSD    lossLOCAL        
 Min.   :      0   Min.   :     0   Min.   :0   Min.   :        0  
 1st Qu.:      0   1st Qu.:     0   1st Qu.:0   1st Qu.:        0  
 Median :      0   Median :     0   Median :0   Median :        0  
 Mean   :  24216   Mean   :  7770   Mean   :0   Mean   :   345969  
 3rd Qu.:    564   3rd Qu.:  1000   3rd Qu.:0   3rd Qu.:     1400  
 Max.   :4515837   Max.   :522569   Max.   :0   Max.   :133500220  

   cropdamage      agriculture         power       relief 
 Min.   :     0   Min.   : 0.000   Min.   :0   Min.   :0  
 1st Qu.:     0   1st Qu.: 0.000   1st Qu.:0   1st Qu.:0  
 Median :     0   Median : 0.000   Median :0   Median :0  
 Mean   :  2769   Mean   : 3.156   Mean   :0   Mean   :0  
 3rd Qu.:   284   3rd Qu.: 2.000   3rd Qu.:0   3rd Qu.:0  
 Max.   :186530   Max.   :90.000   Max.   :0   Max.   :0  

     forest        
 Min.   :     0.0  
 1st Qu.:     0.0  
 Median :     0.0  
 Mean   :   667.7  
 3rd Qu.:     0.0  
 Max.   :186022.0

Subsets by Disaster Type

indo.agg$death <- as.numeric(indo.agg$deaths>0)
floods<-subset(indo.agg, Event == "FLOODS", drop=FALSE)
fires<-subset(indo.agg, Event=="FIRE", drop=FALSE)
forestfires<-subset(indo.agg, Event=="FOREST FIRE", drop=FALSE)

Aggregating by Island

indoisland.agg <- indo %>%
  group_by(Event, Island, Year) %>%
  summarise(events=sum(DataCards, na.rm=TRUE), deaths=sum(Deaths, na.rm=TRUE), injured=sum(Injured, na.rm=TRUE), missing=sum(Missing, na.rm=TRUE), housesdestroyed=sum(Houses.Destroyed, na.rm=TRUE), housesdamaged=sum(Houses.Damaged, na.rm=TRUE), victims=sum(Victims, na.rm=TRUE), affected=sum(Affected, na.rm=TRUE), evacuated=sum(Evacuated, na.rm=TRUE), lossUSD=sum(Losses..USD, na.rm=TRUE), lossLOCAL=sum(Losses..Local, na.rm=TRUE), cropdamage=sum(Damages.in.crops.Ha., na.rm=TRUE), agriculture=sum(Agriculture, na.rm=TRUE), power=sum(Power.and.Energy, na.rm=TRUE), relief=sum(Relief, na.rm=TRUE), forest=sum(Plantation.Forest, na.rm=TRUE))

Data by Island

head(indoisland.agg)

Source: local data frame [6 x 19]
Groups: Event, Island [1]

   Event Island  Year events deaths injured missing housesdestroyed
  (fctr) (fctr) (int)  (int)  (int)   (int)   (int)           (int)
1   FIRE   Java  1998      1      0       0       0             362
2   FIRE   Java  1999      1      0       0       0              38
3   FIRE   Java  2001     21      0       3       0              18
4   FIRE   Java  2002     16      0       0       0            1996
5   FIRE   Java  2003     45     23     791       0            3370
6   FIRE   Java  2004     56     18       1       0            1971
Variables not shown: housesdamaged (int), victims (int), affected (int),
  evacuated (int), lossUSD (int), lossLOCAL (dbl), cropdamage (dbl),
  agriculture (int), power (int), relief (int), forest (int)

summary(indoisland.agg)

         Event           Island        Year          events      
 FIRE       :56   Java      :35   Min.   :1990   Min.   :  1.00  
 FLOODS     :62   Kalimantan:35   1st Qu.:2002   1st Qu.:  5.00  
 FOREST FIRE:19   Sulawesi  :29   Median :2005   Median : 15.00  
                  Sumatra   :38   Mean   :2005   Mean   : 44.46  
                                  3rd Qu.:2009   3rd Qu.: 56.00  
                                  Max.   :2012   Max.   :525.00  
     deaths          injured         missing        housesdestroyed
 Min.   :  0.00   Min.   :    0   Min.   :  0.000   Min.   :    0  
 1st Qu.:  0.00   1st Qu.:    0   1st Qu.:  0.000   1st Qu.:    0  
 Median :  4.00   Median :    0   Median :  0.000   Median :   52  
 Mean   : 18.11   Mean   : 1485   Mean   :  4.365   Mean   : 1005  
 3rd Qu.: 18.00   3rd Qu.:   27   3rd Qu.:  0.000   3rd Qu.:  578  
 Max.   :347.00   Max.   :60423   Max.   :150.000   Max.   :27570  
 housesdamaged      victims     affected         evacuated      
 Min.   :    0   Min.   :0   Min.   :      0   Min.   :      0  
 1st Qu.:    0   1st Qu.:0   1st Qu.:      0   1st Qu.:      0  
 Median :    0   Median :0   Median :    367   Median :    400  
 Mean   : 1230   Mean   :0   Mean   :  87142   Mean   :  27962  
 3rd Qu.:   51   3rd Qu.:0   3rd Qu.:   5879   3rd Qu.:   5389  
 Max.   :36764   Max.   :0   Max.   :4558659   Max.   :1084376  
    lossUSD    lossLOCAL           cropdamage      agriculture    
 Min.   :0   Min.   :        0   Min.   :     0   Min.   :  0.00  
 1st Qu.:0   1st Qu.:       23   1st Qu.:     0   1st Qu.:  0.00  
 Median :0   Median :     2291   Median :     0   Median :  0.00  
 Mean   :0   Mean   :  1244983   Mean   :  9964   Mean   : 11.36  
 3rd Qu.:0   3rd Qu.:    21853   3rd Qu.:  1787   3rd Qu.:  4.00  
 Max.   :0   Max.   :133501384   Max.   :186718   Max.   :239.00  
     power       relief      forest      
 Min.   :0   Min.   :0   Min.   :     0  
 1st Qu.:0   1st Qu.:0   1st Qu.:     0  
 Median :0   Median :0   Median :     0  
 Mean   :0   Mean   :0   Mean   :  2403  
 3rd Qu.:0   3rd Qu.:0   3rd Qu.:    50  
 Max.   :0   Max.   :0   Max.   :188216

Data by Province

head(indo.agg)

Source: local data frame [6 x 21]
Groups: Event, Province, Year [6]

   Event        Province  Year  Island events deaths injured missing
  (fctr)          (fctr) (int)  (fctr)  (int)  (int)   (int)   (int)
1   FIRE BANGKA-BELITUNG  2008 Sumatra     29      0       0       0
2   FIRE          BANTEN  2003    Java      1      0       0       0
3   FIRE          BANTEN  2004    Java      2      0       0       0
4   FIRE          BANTEN  2005    Java      1      0       0       0
5   FIRE          BANTEN  2009    Java      1      0       0       0
6   FIRE          BANTEN  2010    Java      4      2       0       0
Variables not shown: housesdestroyed (int), housesdamaged (int), victims
  (int), affected (int), evacuated (int), lossUSD (int), lossLOCAL (dbl),
  cropdamage (dbl), agriculture (int), power (int), relief (int), forest
  (int), death (dbl)

summary(indo.agg)

         Event                   Province        Year     
 FIRE       :162   KALIMANTAN TIMUR  : 31   Min.   :1990  
 FLOODS     :290   JAWA TENGAH       : 28   1st Qu.:2003  
 FOREST FIRE: 41   SULAWESI SELATAN  : 27   Median :2006  
                   JAWA BARAT        : 26   Mean   :2006  
                   SUMATERA BARAT    : 26   3rd Qu.:2010  
                   KALIMANTAN SELATAN: 25   Max.   :2012  
                   (Other)           :330                 
        Island        events           deaths           injured       
 Java      :129   Min.   :  1.00   Min.   :  0.000   Min.   :    0.0  
 Kalimantan: 98   1st Qu.:  2.00   1st Qu.:  0.000   1st Qu.:    0.0  
 Sulawesi  : 90   Median :  5.00   Median :  0.000   Median :    0.0  
 Sumatra   :176   Mean   : 12.35   Mean   :  5.032   Mean   :  412.6  
                  3rd Qu.: 12.00   3rd Qu.:  4.000   3rd Qu.:    3.0  
                  Max.   :183.00   Max.   :347.000   Max.   :60149.0  

    missing        housesdestroyed   housesdamaged        victims 
 Min.   :  0.000   Min.   :    0.0   Min.   :    0.0   Min.   :0  
 1st Qu.:  0.000   1st Qu.:    0.0   1st Qu.:    0.0   1st Qu.:0  
 Median :  0.000   Median :    0.0   Median :    0.0   Median :0  
 Mean   :  1.213   Mean   :  279.2   Mean   :  341.7   Mean   :0  
 3rd Qu.:  0.000   3rd Qu.:   77.0   3rd Qu.:    0.0   3rd Qu.:0  
 Max.   :140.000   Max.   :21924.0   Max.   :26623.0   Max.   :0  

    affected         evacuated         lossUSD    lossLOCAL        
 Min.   :      0   Min.   :     0   Min.   :0   Min.   :        0  
 1st Qu.:      0   1st Qu.:     0   1st Qu.:0   1st Qu.:        0  
 Median :      0   Median :     0   Median :0   Median :        0  
 Mean   :  24216   Mean   :  7770   Mean   :0   Mean   :   345969  
 3rd Qu.:    564   3rd Qu.:  1000   3rd Qu.:0   3rd Qu.:     1400  
 Max.   :4515837   Max.   :522569   Max.   :0   Max.   :133500220  

   cropdamage      agriculture         power       relief 
 Min.   :     0   Min.   : 0.000   Min.   :0   Min.   :0  
 1st Qu.:     0   1st Qu.: 0.000   1st Qu.:0   1st Qu.:0  
 Median :     0   Median : 0.000   Median :0   Median :0  
 Mean   :  2769   Mean   : 3.156   Mean   :0   Mean   :0  
 3rd Qu.:   284   3rd Qu.: 2.000   3rd Qu.:0   3rd Qu.:0  
 Max.   :186530   Max.   :90.000   Max.   :0   Max.   :0  

     forest             death       
 Min.   :     0.0   Min.   :0.0000  
 1st Qu.:     0.0   1st Qu.:0.0000  
 Median :     0.0   Median :0.0000  
 Mean   :   667.7   Mean   :0.4665  
 3rd Qu.:     0.0   3rd Qu.:1.0000  
 Max.   :186022.0   Max.   :1.0000

Indonesian Fires & Floods

plot of chunk unnamed-chunk-8

  • Increase in events over time

Indonesian Fires & Floods by Island

plot of chunk unnamed-chunk-9

  • Increase in events over time

Floods by Island

  • 162 floods on average per province
  • Highest increase on Java

plot of chunk unnamed-chunk-10

Fires by Island

  • Increase in events over time
  • Increase in fires in Kalimantan

plot of chunk unnamed-chunk-11

Forest Fires by Island

  • Increase in events over time
  • Increase in fires in Kalimantan

plot of chunk unnamed-chunk-12

Houses Damaged in Fires and Floods

  • Increase in events over time

plot of chunk unnamed-chunk-13

Distribution of Local Loss by Number of Events

plot of chunk unnamed-chunk-14

  • Certain provinces incur more loss from events overall
  • Cumulative number of low loss events
  • Impact of few events with high economic loss

Crop Damages by Province over Time

plot of chunk unnamed-chunk-15

  • Certain provinces with more damage; Increase in floods since 2000

Crop Damages by Province, Events

plot of chunk unnamed-chunk-16

  • Forest fires incure high costs from damage
  • Gorontalo, Sulawesi - increase in floods and fires and number of hectares damaged
  • Aceh and Riau, Sumatra - damage from fires in north, damage from floods in south

Crop Damages by Island, Events

plot of chunk unnamed-chunk-17

Data Analysis

  • Models
lmer(formula = lossLOCAL ~ events + housesdamaged + (1 | death), 
    data = floods)
              coef.est  coef.se  
(Intercept)    70506.47 123704.71
events          4238.39   4699.44
housesdamaged    -10.64     41.17

Error terms:
 Groups   Name        Std.Dev.  
 death    (Intercept)       0.20
 Residual             1766464.21
---
number of obs: 290, groups: death, 2
AIC = 9120, DIC = 9215.9
deviance = 9163.0

\( y_{i}=\beta_0+\beta_{1}x_{i}+\beta_{2}x_{i}+\beta_{3}x_{i}+\epsilon_{i} \)

  • \( y_{i} \) = local economic loss in Rupiah
  • \( \beta_{1} \) = events
  • \( \beta_{2} \) = houses damaged
  • \( \beta_{3} \) = any death (binary)

Data Analysis

  • Models
glm(formula = lossLOCAL ~ events + housesdestroyed + deaths, 
    data = floods)
                coef.est  coef.se  
(Intercept)      75768.72 126330.57
events            4327.04   4727.37
housesdestroyed     -9.43     57.71
deaths           -1195.36   4187.71
---
  n = 290, k = 4
  residual deviance = 8.954042e+14, null deviance = 8.982703e+14 (difference = 2.866181e+12)
  overdispersion parameter = 3.130784e+12
  residual sd is sqrt(overdispersion) = 1769402.09

\( y_{i}=\beta_0+\beta_{1}x_{i}+\beta_{2}x_{i}+\beta_{3}x_{i}+\epsilon_{i} \)

  • \( y_{i} \) = local economic loss in Rupiah
  • \( \beta_{1} \) = events
  • \( \beta_{2} \) = houses destroyed
  • \( \beta_{3} \) = deaths

Discussion

  • Focus on prevention and anticipatory plans
  • Urban evacuation routes
  • Reduce intentional fires

Adaptation Efforts

  • Sea Wall
  • Early Warning System
  • Dredging
  • GM crops

Great Garuda Sea Wall

greatgaruda

  • Prevents storm surges from reaching land
  • Erodes islands, reefs
  • Marine ecosystem damage
  • Doesn't address groundwater extraction problem

Early Warning System

NINO4fires

  • Multi-hazard system
  • Model for recovery strategies, not prevention strategies
  • Columbia Institute, 2008
  • NINO4 Index

Dredging

  • Port of Jakarta
  • Training Dikes lower local sea level depth
  • Sohei Matsuno, Professor
  • Increase of 1.5 meters in 2020 above 2013 level

Genetic Modification

  • Increase of GMO crops: palm oil and bananas
  • High toleration of saline and high water levels

Policy Recommendation

  • Implentation of adaptation efforts
  • Implementation of multi-hazard early warning systems
  • Mitigate effects of Great Garuda with water purification & wastewater treatment
  • Create local community adaptation task forces in provinces

THE END

JAKARTA_FLOOD

Thank you